Saturday, March 2, 2013

Credit Bubbles Again

Consider a speculative bubble in housing.

Various government policies are implemented that increase the demand for housing.    The price rises rapidly.   Those currently owning homes gain, and those heavily invested in homes gain even more.    Rather than simply regretting not having been more invested in housing before the rapid price appreciation, some foolish investors purchase homes in  hope of future capital gains.   They myopically project past price increases into the future.   Other things being equal, this foolish behavior further increases the demand for homes and raises home prices.   The foolish investors generate self-fulfilling expectations.

Other investors, noticing and anticipating this foolish behavior, respond not by selling houses at the temporarily high prices, but buy as well, planning to sell out to a "greater fool," just before home prices start to fall back from what are becoming excessively elevated levels.   Call these "clever" investors, "greater fool speculators."

Of course, some conservative investors might sell at what they see as temporarily high prices.    If there enough such conservative investors, the bubble will never get started.   The foolish investors will buy houses  hoping that high price appreciation from the past will continue into the future, but conservative investors will sell all the houses that the foolish investors want to buy at unchanged prices.  (Or ideally, at prices that appreciate at a rate consistent with the fundamentals.)   When the foolish investors are disappointed, they sell the houses they bought, but the conservative investors are there to buy them back so that the prices never fall below the fundamental values.

But let's pass on the happy scenario where there are sufficient conservative investors who understand the fundamentals to keep home prices on track.     The fools bid up the price of houses and "greater fool" speculators also bid up prices, riding the bubble up, planning to fleece the fools by selling near the peak.

It would be possible for those speculating in housing (either the fools or those hoping to sell to greater fools) to fund their purchases solely with cash.   Perhaps they sell stocks and bonds and purchase the houses.    In terms of asset markets, this would lower stock and bond prices.    This would tend to tend to reduce new stock and bond issues to fund the purchase of capital goods.  

While the decrease in the value of stocks and bonds would tend to reduce consumption, the rising prices of existing homes would tend to raise it.   Still, the higher yield on all of these assets, both capital gains on houses, higher bond yields, and higher yield on stocks might motivate increased saving and reduced consumption.   

Adding international considerations, foreign investors might buy up some of the stocks and bonds, dampening the decrease in the prices of those financial assets.  However, this would tend to raise the foreign exchange rate, encouraging imports of consumer and capital goods while discouraging exports.   

Suppose the foolish investors and "greater fool" speculators can borrow some or even all of the funds needed to purchase the homes.    It is hard to imagine that an opportunity to borrow money to fund the speculation would not result in more home purchases than if the foolish investors and "greater fool" speculators had to pay cash.

If the banks (or other lenders) were very conservative, and limited loans to only a fraction of the fundamental value of the homes, they would contribute little or nothing to the bubble.    However, if banks followed the rule of thumb of requiring a down payment equal to a faction of the purchase price, then the bubble is being at least partly financed by credit.  

If banks were to reduce the down payment percent in anticipation of rapid appreciation of home prices, they would be acting as foolish investors.    Given conventional mortgage contracts, there is only limited room for banks to be clever, greater fool investors.   Of course, if the bubble lasts long enough, it is possible that enough equity would be built up in the homes due to repayment of principal and appreciation in the fundamental value of the homes that a bank would avoid loss.

Suppose rather than banks holding mortgage loans to maturity, they sell them off to government sponsored enterprises that are financed by bonds with an implicit government guarantee.   Or, perhaps mortgages are pooled and shares in the pool are sold off to private investors.    Perhaps some investors purchase such securities believing they are very safe because rapid home appreciation makes the underlying collateral worth much more than the mortgages.   In other words, perhaps there are some foolish investors in mortgage-backed securities.   More interestingly, if these securities are liquid, perhaps some investors may know that home prices are unrealistically high, but plan to sell off their mortgage backed securities just before the peak.   In other words, perhaps there are some "greater fool" speculators lending into the speculative bubble in housing.

Perhaps with some of these assumptions, it is reasonable to describe the speculative bubble in housing as being fueled by credit.     However, the banks must fund the added mortgage loans.   If banks hold the mortgages, then they would be funding the mortgage loans either by reducing other lending or else by attracting additional  deposits.    If other loans are reduced, then the interest rate that can be charged on the remaining loans is higher.   If additional deposits are to be attracted, then higher interest rates must be paid, raising the banks' cost of funds.   

If the banks sell off their mortgages, then the government sponsored enterprises purchasing them must sell more bonds, which increases their funding costs.   If mortgage backed securities are sold, then the yields on those must increase to attract the additional funds away from other uses--other types of loans, equity finance of investment, or else current consumption.

While the housing bubble might be fueled by credit, for the most part, this would be a shift of credit into housing and away from other sectors of the economy.   It is, of course, possible that in increase in interest rates would generate more saving.   That added saving could generate more credit supply.   It also entails less consumption demand.    It is also possible that households might provide less equity finance (hold less of their wealth in stocks) and so the the housing boom would partly be financed by additional credit at the expense of equity finance in other parts of the economy.

More fundamentally, the housing bubble is an increase in investment demand.   It raises the natural interest rate.   The increase in the interest rate raises the amount saved and reduces the quantity of investment demanded.   There is a reduction in spending on consumer goods and services and other types of capital goods.   There is an increase in spending on new homes.

Taking into account international considerations, some of the decrease in consumption and investment might occur in other countries.   They would help fund the speculative bubble in housing.  The net capital inflow would result in a higher exchange rate and so expanded imports and reduced exports.   The reduced exports is a lower demand for output, and the increased imports come at the expense of demand for import competing products.

If the net capital inflow takes the form of foreigners accumulating bank deposits or mortgage backed securities, then these foreign creditors are directly funding the boom.   However, it would be possible for the foreign investors to purchase something else, such as government debt, while domestic investors shift from the government debt in order to hold bank deposits or mortgage backed securities.   Again, it would be reasonable to describe the bubble as being fueled by credit.

There is no increase in nominal GDP.   Instead, there is a shift into spending on new houses and away from spending on other capital goods or consumer goods.   To the degree that occurs in other countries, the reduction in spending on domestic output that matches the increase spending on houses would be reduced spending on import competing goods and export goods.

If the added demand for credit is funded by additional checkable deposits, it would be entirely appropriate that both the yield on money and the quantity of money rise.   On the other hand, to the degree that higher interest rates result in a reduced demand for hand-to-hand currency or other noninterest bearing forms of base money, the quantities of those types of money would need to decrease.

So far, the assumption has been that the increase in credit demand to fund the purchase of added homes results in an appropriate, market clearing increase in interest rates on both home loans and on the credit instruments used to fund the loans.   All that has happened is that from our assumed "God's eye" perspective, house prices have been driven too high because there are not enough conservative investors to offset foolish investors and greater fool speculators.   There has been an increase in saving and decrease in investment on other types of capital goods.   There has been a reallocation of spending towards new homes and away from spending on other goods and services.

Market Monetarists reject the notion that it is desirable to have a central bank that will create a shortage of base money and force interest rates higher than the natural interest rate so that fewer loans will be made to purchase new homes.   Our view is that while a central bank can cause a decrease in spending on housing by causing a shortage of money and raising the market interest rate above the natural interest rate, that will also reduce spending on everything else.   By assumption, spending on everything else is already too low.   If spending on houses is really too high, what should happen is that spending on houses should decrease and spending on everything else should increase.   Creating a shortage of money or raising the interest rate above the natural interest rate would have the exact wrong effect on the other sectors of the economy.    

The Market Monetarist approach is that monetary authority should adjust the quantity of base money to match the demand for base money so that market rates remain equal to the natural interest rate and any increase in spending in one sector of the economy, whether funded out of current income, equity, or credit, should be matched by decreases in spending  in other areas of the economy.   Nominal GDP should remain on a steady growth path, and not be shifted about because spending is "too high" in one sector of the economy.

But what happens when the bubble bursts?    Once the foolish investors recognize reality and there is no more opportunity for the "greater fool" speculators to fleece them, the demand for investment falls.   This implies a decrease in the natural interest rate.   The lower interest rate should result in a decrease in the quantity of saving supplied and an increase in the quantity of investment demanded.     Thankfully, less will be spent on new homes, but lower interest rates will result in more being spent on consumer goods and other capital goods.

With the extra construction of new homes during the boom, there may be a period where new home construction must fall below long run levels.  (In a static world, that would require waiting until enough of the existing homes have depreciated.   If the demand for new homes is usually growing, then a pause or slowdown in new construction will soon bring the supply of homes to a level consistent with now larger demand.   In a growing economy, the boom didn't result in too many houses, exactly, but resulted in homes that will be needed eventually being produced too soon.)  

Of course, the reduction in the production of other capital goods during the boom also created a "hangover" in the other direction.     Too few of them were produced, and there are opportunities to catch up.  This would make investment demand more elastic with respect to the interest rate in the short run, and so result in a smaller decrease in the natural interest rate.

On the other hand, to the degree that increases in home prices had led to reduced saving and increased consumption by homeowners during the boom, then once home prices return to their fundamental values, it might be necessary for homeowners to increase their saving and rebuild their wealth.   (If anyone had been saving more because of the higher interest rates during the boom, then they might have an offsetting bulge of unsatisfied consumer demand.)   This addition to saving would imply that the immediate decrease in the natural interest rate would be larger, and while there would be a decrease in the quantity of saving supplied due to the lower natural interest rate, the net effect of this desire to rebuild wealth would be more spending on capital goods (other than new homes,) rebuilding wealth.

However, as above, the added consumption due to the added wealth of homeowners during the boom is balanced by reduced wealth of those owning other financial assets, due to the higher interest rates.  Once the boom collapses, and housing prices fall, that reduced wealth is balanced by an increase in stock and other assets prices due to the lower natural interest rate.

What about international considerations?    Much of the added investment could be in other countries.   The lower interest rate would tend to result in a decrease in the exchange rate, resulting in additional spending on export goods (perhaps capital goods for foreign markets.)   Also, the reduced exchange rate would raise import prices and so result in added demand for domestic import-competing industries.

There should be no reduction in nominal GDP.    Spending on new houses falls, while spending on other capital goods and consumer goods rises.    The added spending may be in other countries directly, and so maintaining domestic production might require added exports and increased demand for import competing industries.

What about money?  

To the degree the lower natural interest rate results in lower interest rates on checkable deposits, then it is possible that the quantity of checkable deposits and so, the quantity of money, should fall.   However, with other interest rates falling as well, this would really just be an aspect of a possible reduction in the quantity of saving supplied and increase in consumption demand.    On the other hand, to the degree that these lower interest rates increase the demand for base money, the quantity of base money should rise.

What about finance?  

Of course, the demand for home loans falls when the bubble pops. With the reduction in credit demand, the banks would need to lower the interest rates they charge on loans and what they pay on deposits.   The lower interest rates on consumer loans would increase demand for consumer goods, particularly consumer durables.  Lower interest rates on commerical and industrial loans would result in increased demand for capital goods.

 The lower interest rates on deposits might motivate those holding deposits to reduce saving and purchase consumer goods.   Howerve, they might be motivated to instead purchase stocks or bonds.   

The firms purchasing capital goods issue new stocks and bonds.   Of course, many might use current revenue to purchase capital goods or even sell off  prevously accumulated bank deposits or stocks and bonds.

To the degree that there has been a large financial infrastructure built for boom time financing of homes, the end of the bubble will make individuals and firms specialized in that area substantially worse off.   For example, brokers in mortgage backed securities might have little to do.   

More troubling is the possibility that many investment and commerical banks might suffer such heavy losses on their portfolios of mortgages (and mortgage backed securities) that their ability to provide services would be compromised.   While that aspect of their business should be curtailed for a time, there are opportunities to underwrite new stocks and bonds and make additional consumer, commercial, and industrial loans.

Of course, to the degree there remain sound financial institutions, it is those that should expand their underwriting and lending activity.   The shift in market share away from those that lent into a speculative bubble towards those that avoided the error would provide the correct incentives to avoid such errors in the future.    Where would these "sound banks" obtain the funding to greatly expand lending?    Presumably from those who are no longer despositing their funds in unsound banks.

Maintaining overall equilibrium, understood as keeping the market rate equal to the natural interest rate, the quantity of money equal to the demand to hold money, and spending on output growing on a stable growth path, could well require a period of very low interest rates on bank deposits.   Lower interest rates on checkable deposits reduces the demand to hold money and makes it more attractive for banks to fund assets by issuing money--checkable deposits.    This will reduce the banks' funding costs and so allow them to lower lending rates in a way consistent with the reduction in the natural interest rate.   Finally, it will enhance the profits of the sound banks that should be expanding while curtailing the losses of the less sound banks whose contraction should be dampened in line with the ability of sound banks to expand.

If there is a very large segment of the banking system where losses have made it difficult to expand what are now profitable lines of business, a rapid reorganization might be in order.   In the limit, where every single bank is absolutely insolvent because of past bad loans, then such a rapid reorganization is essential.   Closing down every single bank and gradually liquidating them, while allowing new banks to be opened, is probably not the most sensible approach to such an extreme situation.   (Kevin Dowd once likened it to rolling the patients of an insolvent hospital onto the sidewalk.)

To the degree that financial intermediation is greatly hampered, then much, and in the limit, all, of the expansion in investment would need to occur by direct finance.   Firms enjoying increased sales would need to use their added cash flow to purchase additional capital goods.   Households that would otherwise be inclined to save out of current income would need to instead purchase existing stock in the profitable firms expanding by retained earnings.  However, if those stock prices rise too high, the sensible response would be for such households to expand consumption spending, perhaps concentrating on durable consume goods.   Further, households that are collecting on loans that are being repaid might need to use those funds to purchase stock or even "dissave," again purchasing consumer durables.

Clearly, a banking system that is impaired in this way, with the healthy parts expanding rapidly but there being a large remainder unable to operate, is worse than a banking system where each and every part is operating optimally.   In the context of a fundamental reduction in investment demand and possible increase in saving supply, the greater impact of hampered intermediation needs to be lower interest rates paid on deposits of all sorts.   Most importantly, the interest rates on checkable deposits needs to be low enough to keep the demand to hold them in line with the quantity the banking system is able to issue in its impaired state.  

The analysis of the speculative bubble in housing and its end is in the context of a nominal GDP target.   The quantity of money adjusts to the demand to hold it and the market rate adjusts with the natural interest rate in the context of a stable growth path for nominal GDP.

Suppose the analysis is done again with the assumption that a central bank is holding interest rates constant the entire time so that the central bank creates an excess supply of money and keeps the market rate below the  natural rate in the boom and then creates an excess demand for money and keeps the market rate  above the natural interest rate when the bubble pops?  

Suppose the central bank doesn't keep its policy interest rate fixed, but is always a step behind?

Suppose the central bank targets the exchange rate so that when the exchange rate "should" rise in the boom, the central bank creates an excess supply of money and keeps the market rate below the natural rate?   And then, when the exchange rate "should" fall after the bubble pops, it creates an excess demand for money and keeps the market rate above the natural interest rate?

Suppose the analysis is done such that the central bank targets inflation?   Worse, suppose it only looks at consumer prices?  Still worse, suppose it uses rent equivalent to measure the "price" of housing?    With a speculative bubble, the added production of new homes reduces rents.    It is only as added home production pulls resources away from the production of other consumer goods and services that resource prices, including wages, result in higher prices of other consumer goods.

When the bubble pops, the rapid fall off in spending on new houses should be offset by added spending on other goods and services.     The resources no longer used to produce houses can and should be used to produce other useful goods and services.   But a central bank that looks only at house rentals would ignore the direct effect off the drop in the prices and quantities of new homes.   The reduced production of new homes tends to increase rents.   It is only as a growing recession leads to reduced demand for other consumer goods and services (and rents on homes) that consumer prices would fall (or grow more slowly.)  

With nominal GDPlevel  targeting, if spending on new houses grows faster, and spending on other sorts of goods and services do not grow more slowly, then nominal GDP will rise above target.   With level targeting, this will require slower growth of nominal GDP in the future, and that expectation will directly slow spending on various goods and services now, slightly dampening the underlying "housing bubble" while slowing the rest of the economy to provide the resources needed to produce these additional houses.  

When the bubble pops, and spending on new homes falls, then if spending on other goods and services does not rise, then nominal GDP will fall below target.   If that happens, then nominal GDP must grow more quickly to catch up to the target.   This will slightly offset the decrease in demand for housing, but more importantly, stimulate spending on other goods and services.

Again, if the consequences of the bubble popping are so catastrophic that despite a regime of nominal GDP level targeting, the equilibrium nominal interest rate on checkable deposits falls below the cost of storing base money currency, then there is a problem.    Market Monetarists have a variety of solutions to the problem of the "zero nominal bound," starting with not reifying the policy interest rate so that when it hits zero, there becomes a wrongheaded notion that nothing more can be done.   Avoiding a "growth rate" target,  like inflation, which exacerbates the problem, is a close second.

Causing recessions by creating an artificial shortage of money and pushing the market rate above the natural interest rate in order to stop what the central bank decides is to much spending in one sector, even if it is being financed with credit, is not a sensible solution to the possible problem of a zero nominal bound.

1 comment:

  1. Bill
    This post "charts the argument" against "credit bubbles"